Thread, while I don't reply to every instance of contention or
misinformation being presented on this thread, with regards
DSL and Cable Modem properties and attributes, I would like to
clarify a few things that have popped off from my screen that
cry out for moderation, and in some cases, triage.
-----
Comparing HFC to GBLX. This may seem trite, but since
someone mentioned it, it demonstrates a need to contrast the
nature of HFCs to those of GBLX, and other f-o carriers'
systems. Cable TV HFCs cannot be likened to those of the
fiber barons'. Perhaps they should be at some time in the
future, hopefully soon, but unfortunately this is not the case
today.
The former (HFC) entails the delivery of RF carrier
supported information (TV signals and cable modem
information that has been modulated) on fiber between the
head end location (central site) to a field node closer to
subscribers. There are may variations as to the specifics.
At the field node the RF is down modulated to lower
frequencies and sent out to the residence over the coaxial
line. At the terminal point it goes to an STB, ordinarily, and
then to the cable modem.
GBLX's fiber construction and the transport scheme that sits
atop it is meant only to transport ultra high bit rate digital
information via SONET/SDH and future photonic flows.
These futures may be more of the same plus some newer,
modified SONET/SDH, and increasingly over time,
intrinsically lambda-in-nature (approaching purer forms of all
optical) transport and routing modes.
----
The matter of security over HFC comes up frequently, and
requires some care, as well. When was the last time anyone
here has heard of a substantiated horror story of a cable
TV-supported breach of security in the press? Cable's
antagonists would seize upon such opportunities jumping all
over them, exploiting them for maximum effect. I don't hear
any of this at this time, do you?
The reality is that very few breaches are ever committed on
HFCs, and if they are, the perps are incredibly secretive
about it, which is extremely uncharacteristic of that breed.
And when these security infractions are committed,
they are not by someone sitting on a couch with a can of
beer in one hand, with the other one tuning into the cable
sphere for amusement or entertainment purposes, although
I'm certain that this has been attempted to no avail by many.
It's usually someone with special hacking skills aided by
some relatively esoteric test equipment not yet found at the
corner Radio Shack, and who possess more than just a little
knowledge of link level security features which are now
being used. An individual with these skills could probably
find a greater catch on the open 'net, itself.
And if that weren't enough for the time being, Cable Labs
has provided for higher levels of protection within the most
recent release and future planned releases of the cable
modem DOCSIS standard.
Granted, during the early test and pilot days of exploring
cable's potential, there were indeed some horror story class
penetrations that took place. Horrendously violative, and
humorous at the same time... depending on who you are,
and if it affected you ;-)
I could recall one such event over in Bell Labs that has
reached legendary proportions by now, but if I told you how
that one was facilitated (i.e., what type of breadboard level
network construction was used) you would realize that the
breadboard system it was committed on bore no
resemblance to today's systems.
Things have changed since then in this respect, dramatically,
and they will continue to become even more secure, going
forward.
----
"Will Terayon's S-CDMA scheme (and while we're at it,
CMTO's ATM capabilities) aid security?" was one of the
more stimulating questions which I saw presented upstream.
My personal opinion is that they will. In the respect that
TERN's, just as CMTO's, approach would require the
hacker to be even more astute in their evil, voyeuristic ways.
Cracking both the link level authentication, as well as the
elusive spread spectrum scheme of TERN, and the ATM
bit-scrambling technique now used in CMTO, would just
add to the hacker's level of achievement if they were
successful. However, these additional obstacles make such
an achievement even more unlikely. And if these aren't
enough, there's the additional DOCSIS encryption option
that would need to be negotiated soon, as well.
----
Relating to the symmetrical capabilities or lack thereof of
cable, most cable systems, like DSL, are asymmetric in
nature due to the paltry and hostile provisions made
available on the lower end of the spectrum where the
upstream is supported. Newer modulation and transport
schemes, however, again, citing TERN's and CMTO here,
actually do allow for same-speed up and down without the
need for adding more coax, as someone has suggested.
While the upstream still presents a formidable problem to
each of these that I've mentioned, and despite the somewhat
reduced overall speeds that they achieve as a penalty that
partially offsets their gains, symmetry is nonetheless now
possible using these makes. Perhaps others here can name a
other makes that can also support up and down symmetry,
as well.
----
As for the rest of the sometimes extremely informed
discussion that takes place here on this topic, amidst some
noise at times that must be checked and corrected, I'll leave
that to you gladiators of the ether who are more than
capable in fending for yourselves. In this respect I've got
more than I can contend with right now over in the Last Mile
and ATHM threads. I just wanted to make these
clarifications.
Disclaimer: I do not own or trade any of the stocks
mentioned in this post, nor do I own or trade any other
individual companies in this sector.
As a request, I'd appreciate it if Dr. Levy, and others
similarly qualified, did a critique on this post, especially on
my takes concerning TERN and CMTO.
Dave Horne, say something here, if you will, about security.
George, other issues which I'd like to address in another post
perhaps, sometime later during the weekend, are the issues of DSL's
suitability in certain networking venues that were mentioned here,
such as its comparative merits where speed and its role in multi-
casting, are concerned.
Best Regards to All,
Frank Coluccio |
